wrapMatrixEQTL: QTL Analysis for expression and methylation data.

Description Usage Arguments Details Value Author(s)

View source: R/SYB_wrapMatrixEQTL.R

Description

wrapMatrixEQTL uses MatrixEQTL package for QTL analysis of expression or methylation data with SNP genotypes.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
wrapMatrixEQTL(inputset, SNPfile_transpAddCoded_name, SNPfile_tfam = NULL,
  covariates_file_name = NULL, covarsampleID = "IID", covar2adjust = NULL,
  projectfolder = "QTL", projectname = NULL, sampleColumn = "Sample_Name",
  useModel = modelLINEAR, errorCovariance = numeric(), QTLtype = "both",
  cisDist = 1e+06, pvOutputThreshold_cis = 1e-06,
  pvOutputThreshold_tra = 1e-06, pvOutputThreshold_all = 1e-06,
  genepos = NULL, genemart = useMart("ENSEMBL_MART_ENSEMBL", host =
  "feb2014.archive.ensembl.org", dataset = "hsapiens_gene_ensembl"),
  ensembl_filter = "illumina_humanht_12_v4", updateSNPpos = FALSE,
  snpmart = useMart("ENSEMBL_MART_SNP", host = "feb2014.archive.ensembl.org",
  dataset = "hsapiens_snp"))

Arguments

inputset

Either ExpressionSet, SummarizedExperiment, DESeqDataSet, MethylSet or data.frame containing summarized methylation Island data from COHCAP.avg.by.island-function (package COHCAP).

SNPfile_transpAddCoded_name

character with path to genotype file (transposed and additive coded!).

SNPfile_tfam

character with path to tfam file with sample information. If NULL, sample names are expected as header line of SNPfile_transpAddCoded_name. If given, SNPfile_transpAddCoded_name is read as no-header file and sample names are added from SNPfile_tfam as column names. Check in advance if you need SNPfile_tfam or not.

covariates_file_name

character with path to covariates file.

covarsampleID

Character with column name of sample IDs in covar file.

covar2adjust

Character vector with column names of covariates to adjust QTL analysis. Omitted if NULL.

projectfolder

Character containing path to output folder (will be generated if not existing).

projectname

Character used as suffix for output files.

sampleColumn

Character with column name of sample IDs in input dataset.

useModel

model to use (modelANOVA or modelLINEAR or modelLINEAR_CROSS). Set useModel = modelLINEAR to model the effect of the genotype as additive linear and test for its significance using t-statistic. Set useModel = modelANOVA to treat genotype as a categorical variables and use ANOVA model and test for its significance using F-test. The default number of ANOVA categories is 3. Set otherwise like this: options(MatrixEQTL.ANOVA.categories=4). Set useModel = modelLINEAR_CROSS to add a new term to the model equal to the product of genotype and the last covariate; the significance of this term is then tested using t-statistic.

errorCovariance

Numeric error covariance matrix. Use numeric() for homoskedastic independent errors.

QTLtype

Character with "both" for calculating cis and trans QTL separately, "all" for no separation of cis and trans, "cis" for cis QTLs only.

cisDist

Numeric maximal baisepair distance for cis gene-SNP pairs.

pvOutputThreshold_cis

Numeric significance threshold p-value for cis QTL tests.

pvOutputThreshold_tra

Numeric significance threshold p-value for trans QTL tests.

pvOutputThreshold_all

Numeric significance threshold p-value for all QTL tests (cis and trans mixed).

genepos

dataframe with 4 columns (geneid, chr, left, right) or character with path to gene position file. If NULL, genepos file is generated with biomaRt.

genemart

biomaRt object to be used for updating gene positions.

ensembl_filter

Character with filter name to search in genemart.

updateSNPpos

Boolean. Shall SNP-positions be updated via biomaRt?

snpmart

biomaRt object to be used for updating SNP positions.

Details

Expression or methylation data can be processed for QTL analysis. Coordinates for genes and SNPs are either given in SNPfile_transpAddCoded_name or genepos, respectively, or are downloaded from biomaRt. Model type and optional covariates can be selected for analysis. Analysis can be perfomed for either cis or trans QTLs or both simulatanously.

Value

list containing analysis parameter and QTL results. Intermediary results and plots are stored in projectfolder as side effects.

Author(s)

Frank Ruehle


frankRuehle/systemsbio documentation built on Dec. 8, 2018, 10:27 p.m.